ENTS Points  

TOPIC: Sneak preview

== 1 of 2 ==
Date: Wed, Sep 24 2008 7:25 am
From: dbhguru@comcast.net


Folks, it is time to reconsider our two ENTS methods of ranking the size of trees: ENTSPTS and TDI. The TDI system is sound. No modifications needed there, but ENTSPTS is ailing, the reason being that the number of points awarded does not track well enough with increases in trunk volume . The following table compares the effect of tree size increases using the old way of calculating ENTSPTS ( height x circumference) , a proposed new way of calculating ENTSPTS ( [height x Circumference ^2]/100), and an abbreviated version of the champion tree formula ( 12 x circumference + height).

   Height     Circ  VOL-CONE ratio   ENTSPTS   ratio   ENTSPTS2    ratio        Champ      Tree Pts    ratio
50 8 84.8   400   32   146  
50 12 190.8    2.3 600    1.5 72     2.3 194     1.3
50 16 339.2    4.0 800    2.0 128     4.0 242     1.7
100 8 169.6    2.0 800    2.0 64     2.0 196     1.3
100 12 381.6    4.5 1200    3.0 144     4.5 244     1.7
100 16 678.4    8.0 1600    4.0 256     8.0 292     2.0
150 8 254.4    3.0 1200    3.0 96     3.0 246     1.7
150 12 572.4    6.8 1800    4.5 216     6.8 294     2.0
150 16 1017.6   12.0 2400    6.0 384    12.0 342     2.3

Looking at the table, we see that the ratio of the volume of the largest tree to the volume of the smallest is 12 to 1. The ratio of ENTSPTS of the largest tree to the smallest is 6 to 1. The ratio of modified ENTSPTS of the largest to the smallest tree is 12 to 1 (just what we want), and the ratio of modified champion tree points of the largest to smallest tree is 2.3 to 1. The change in modified ENTSPTS tracks perfectly with conical volume. Each ratio in the above table is the current entry divided by the first entry in the respective column, not the preceding entry in the column. The purpose of the ratio columns is to show how points track with changes in volume as measured by a form such as the cone or paraboloid.
The reason I chose a scaling factor of 100 for modified ENTSPTS is to bring the point total more in line with numbers that come from the champion tree formula. Additionally, it is computationally simple. I leave out hypothetical crown spread in the table. However, were we to include realistic crownspreads for the size trees indicated by height and circumference, the ratio of the points of the largest tree to the smallest would increase slightly - perhaps 2.5 to 1.
I've discussed the new system of ENTSPTS with Ed off list. Ed is solidly behind it. Ed also mentioned that John Eichholz had once before pointed out the value of C^2 versus C as the factor dealing with circumference. I mentioned the proposed new method briefly to Will in a recent phone conversation and told him I'd shortly present some analysis. The above table is the first step in that direction.
Thoughts anyone?

== 2 of 2 ==
Date: Thurs, Sep 25 2008 2:02 am
From: Beth Koebel

Not being a math major (I had to drop CAL I because I couldn't understand it),  it looks like you are using a cone to measure the volume as the "gold standard" and then using the new ENTPTS2 to get the measurements that are often taken, height and circumfence, to match it.  If this is the case, then would this work also for trees like palms or any other tree in which there is a trunk without branches for say 50 or so feet then a relatively flat crown(umbrella shaped)? How about the classic hardwood shaped tree (golf ball on a tee)?
BTW, I am not going to be able to make it to the ENTS gathering in Oct. as it is too close to my projected closing.  Sorry, I wish I could've made it.  Maybe the next one.

== 2 of 11 ==
Date: Thurs, Sep 25 2008 6:06 am
From: dbhguru@comcast.net


The proposed ENTS point formula admittedly works best for trees with long straight trunks that can be modeled with a regular geometrical form, principally a neiloid, cone, or paraboloid. I chose the cone for illustration purposes, but either of the other two forms would have worked just as well.
The question of what kind of formula works for a big spreader like the live oaks that Larry measures is probably not going to be adequately determined for a long time. There is just too much wood tied up in the complex network of limbs. The ENTSPTS formula is not the answer for trees of that shape, but then neither is the champion tree formula. Consider the table below.

    HGT             CIR       SPD      CHP PTS       ENTSPTS
              50              12           120 224 72
65 24 120 383 374.4
130 24 120 448 748.8

For trees with spreads of 120 feet, we know there is lots of wood committed to the limbs. Looking at the entries in the table, it is apparent that ENTSPTS does not capture limb wood. The champion tree formula actually does better, but going from rows 2 to 3 is just not logical for the champion tree formula. A 130-foot tall tree with a 120-foot crownspread implies a lot more wood than the spread of points of 383 to 448 indicates.
The problem we're experiencing in calculating an absolute number of points for a tree stems from the one size fits all approach. I understand that it was for simplicity's sake and to try to get the general public involved, but the formula doesn't work well enough for a group like ENTS.
For a system of relative comparisons, TDI works well and we may never get beyond that, i.e. relative comparisons. However, for white pines in New England, I need more of an absolute measure. The amount of limb mass for a tall, straight conifer may not be more than 5% or 6% of trunk volume. So, I don't have to worry too much about the limbs and can apply the proposed formula. By contrast, the limb volume versus trunk volume ratio may approach 50% for live oaks. I wouldn't apply to formula to trees of those shapes. So, the search must go on.
I apologize to the list for not making it clear that I had conifers in mind for the proposed formula. Very clumsy of me.
Sorry you won't be able to make it to the rendezvous. The one in 2009 will be in Cook Forest. That is considerably closer to help for time and expense travel.


== 3 of 11 ==
Date: Thurs, Sep 25 2008 9:01 am
From: "Edward Forrest Frank"


I think you are equivocating too much. Your proposed formula provides a number that is proportional to the trunk volume of a tree with any regular geometric shape or combination of shapes ranging from neloid, to conical, to parabaloid. It is approximately equivalent to the 1/8 the volume of a cylinder of the same circumference and height. Individual trees even within a species vary quite a bit, so the formula will not replace the volume measurements of an individual tree, but will trend parallel to the volume of trees in general. If a particular tree species tended to fall in a given range, say 30 to 40% of the volume of a cylinder of the same size, then a simple constant - a number- could be multiplied by the ENTS formula to determine what the typical trunk volume or the range of typical trunk volumes would be for a tree of that species for those basic height and girth measurements. I might expect that different tree species may have a tendency toward one typical shape or another, or that trees of a certain age range may may have a tendency toward one shape over another. But in all cases the proposed ENTS points will in general trend in the same direction as trunk volume. As Bob says the formula only deals with the trend of trunk volume and not with the volume of the tree limbs. Trees with broadly flaring bases or tops that are broken off will be outliers to the overall trends also. Since volume is proportional to the square of the circumference for any regular geometric shape, it is certainly a much better rough approximation of tree volume trends than the simple product of height and girth..


== 4 of 11 ==
Date: Thurs, Sep 25 2008 10:14 am
From: dbhguru@comcast.net


I'm content with the formula as a replacement for Cir x Hgt. For trees with long trunks, I think it is definitely preferable to the champion tree formula. We're making progress.